Optimal integration of EGFR inhibitors in advanced colorectal cancer

With two classes of approved targeted agents in colorectal cancer, there was natural interest in combining these agents. Initial promising data came from the BOND-2 randomized phase II trial, in which 83 patients refractory to irinotecan but naïve to EGFR- and VEGF (vascular endothelial growth factor)–targeted agents were randomized to receive cetuximab-bevacizumab with (CBI) or without (CB) the addition of irinotecan.³⁵ Irinotecan was given at the same dose and schedule at which the patient’s disease had previously progressed. The median TTP in the CBI arm was 7.3 months, whereas it was 4.9 months in the CB arm. The response rates were 37% and 20%, respectively. With a median follow-up of 28 months, the OS for the CBI arm was 14.5 months versus 11.4 months without irinotecan. As with BOND-1, this trial suggested that targeted therapy could restore sensitivity to irinotecan, and the encouraging response rates and survival data prompted phase III study.

In the PACCE trial, the addition of panitumumab to bevacizumab-chemotherapy was studied in 1,053 patients previously untreated for metastatic disease.³⁶ They received either oxaliplatin- or irinotecan-based chemotherapy (investigator choice) and then were randomized to receive additional panitumumab or no additional treatment. Almost 70% of patients received oxaliplatin-based chemotherapy. Despite the promise of BOND-2, the trial was halted early, after an interim analysis revealed increased toxicity and inferior PFS in the panitumumab cohorts. The PFS was 10.5 months in the bevacizumab-oxaliplatin–containing arm and 8.8 months with the further addition of panitumumab (HR = 1.44; P = 0.004). In the irinotecan-based cohorts, a similar decrement was seen with the addition of panitumumab (PFS, 11.9 vs 10.1 months; HR = 1.57). Even when the results were retrospectively analyzed by KRAS status, there was no improvement in outcomes seen in the KRAS wild-type population.

The inferior outcome with the combination of EGFR- and VEGF-directed therapy was confirmed in the phase III CAIRO-2 trial, in which 755 previously untreated patients were randomized to receive capecitabine-oxaliplatin-bevacizumab with or without the addition of cetuximab.³⁷ As with the PACCE study, PFS was the primary endpoint here, and initial results were later retrospectively analyzed according to KRAS status.

In the full study population, the addition of cetuximab was associated with a decrease in PFS from 10.7 to 9.4 months (HR = 1.22; P = 0.01). In the KRAS wild-type population, PFS was almost identical (10.6 months vs 10.5 months), and there was no difference in OS (22.4 months vs 21.8 months; P = 0.64). This result suggested that the lack of benefit for combining targeted agents seen in the PACCE study is not restricted to panitumumab but rather applies to the combination of bevacizumab and EGFR-directed monoclonal antibodies as a class.

The use of EGFR-targeted therapy in the neoadjuvant treatment of advanced colorectal cancer may have the potential to increase the resectability rate in patients with isolated liver metastases. This remains an active area of investigation, but several reports have already been published. In the CRYSTAL trial, first-line addition of cetuximab to FOLFIRI increased the response rate in KRAS wild-type patients (39.7% vs 57.3%), and there was a statistically significant increase in the rate of R0 resection with curative intent (1.7 vs 4.8%; P = 0.002).21

The CELIM trial was a randomized phase II study evaluating FOLFOX6-cetuximab versus FOLFIRI-cetuxi¬mab specifically in a population of patients with unresectable liver-only colorectal metastases.¹³ The primary endpoint was response rate, which was quite good in both groups (68% with FOLFOX6-cetuximab, 57% with FOLFIRI-cetuximab). KRAS status was evaluated retrospectively, and in the wild-type population, the combined response rate was 70% versus only 41% in KRAS-mutant patients, a difference that was highly statistically significant (P = 0.008). In this study of patients deemed initially unresectable, R0 resections were subsequently performed in 34% of patients. The use of EGFR antibodies in the neoadjuvant setting is especially appealing, given the consistently high response rates seen and the fact that the largest phase III trial evaluating bevacizumab with chemotherapy in the front-line setting demonstrated no improvement in response rate.¹⁰

There are a number of ongoing clinical trials that will provide important data to help clinicians choose the optimal targeted regimens for their patients. KRAS wild-type status is now a standard inclusion criterion for ongoing trials.